The minimum threshold scan consists of multiple scan steps (AliITSOnlineSPDscanMultiple class, see SPDscan Package documentation). Detector configuration changes throughout the scan, applied in such a way that the chips get lower and lower threshold settings.
The purpose is to find the lowest threshold value (highest efficiency), at which the chip is still operating noise free. Here is a movie from a scan made with SPDMood: minth2.mpg
The minimum threshold values are retrieved from the "AliITSOnlineSPDscanAnalyzer::GetMinTh" method. Graphs of the mean multiplicity vs. scan step are analyzed for each chip.
- Find DAC value just before mean multiplicity exceeds (fMinIncreaseFromBaseLine+[mean multiplicity of step 0]).
- A baseline reference value is estimated from an error function (s-curve) fit to the graph from step 0 to step found in 1. A maximum allowed value is set through fMaxBaseLineLevel.
- Find step closest (lower) to the DAC value in 1.
- The steps 1 to 2/3 of step found in 3 are analyzed. A compare value is estimated 2 sigma above the baseline, or at (fMinIncreaseFromBaseLine+baseline) if this is lower.
- The Minimum Threshold is the DAC value of the highest step, where the mean multiplicity is below the compare value in 4. To be on the safe side, the value is stepped down by fStepDownDacSafe.
The tuning parameters in italic are defined in the file standal_params.txt
|fMinIncreaseFromBaseLine||required increase from the baseline, used to find where the rise begins|
|fMaxBaseLineLevel||maximum allowed baseline level, it should be kept low by masking noisy pixels|
|fStepDownDacSafe||safety margin: the threshold found by the algorithm will be decreased|